The urothelium lines the bladder and protects it from bacteria and toxic or carcinogenic substances in the urine by maintaining an impermeable barrier. Although this functions remarkably well, it fails in bladder cancer, interstitial cystitis, and possibly other diseases as well. Understanding how the urothelium grows and differentiates and carries out its function may provide important clues to the diagnosis and treatment of bladder diseases. Although urothelium can be cultured, most models fail to adequately model the interactions between the urothelium and the underlying stroma. Together the two form a complex and interacting system. This proposal describes studies to meet the requirements of a PAR to develop cell-selective tools, specifically item (8) identification of cell specific markers to aid studies of epithelial-stromal interactions in normal and malignant tissues and secondarily item (5) discovery of biomarkers that indicate health or mass of individual cell types and item (1) discovery of genes selective to individual cell types. Novel methods for growth of urothelium as an artificial tissue in 3-dimensions and for bladder cancers as artificial tumors on both normal and cancer-modified matrix provides a basis for discovering the genes and proteins involved in defining normal and malignant urothelium. In a first discovery phase, microarrays and a novel proteomics approach with 2-D chromatography will be combined with innovative bioinformatics approaches to identify candidate genes and proteins. A public database of gene expression and protein levels will be constructed. Candidate genes/proteins will be validated sequentially, first by data mining and informatics to identify plausible mechanisms and to compare with other expression data, then biochemically by immunohistochemistry or quantitative PCR. A limited number of the most interesting genes will be investigated in greater depth in a second phase of hypothesis testing. Expression will be forced in cell lines not expressing a gene or knocked down in those that do. The "deliverables" will be sets of genes and proteins subjected to several levels of verification for their role in aiding studies of epithelial-stromal interactions in normal and malignant tissues.